Mohammad Idrees Adhi1, Syed Moyn Aly2. 1. King Abdul Aziz Medical City & King Abdullah Specialized Children Hospital, National Guard Health Affairs, Riyadh, Saudi Arabia. 2. Department of Medical Education, Jinnah Sindh Medical University, Rafique Shaheed Road, Karachi, Pakistan.
Abstract
OBJECTIVE: To find differences between One-Correct and One-Best multiple-choice questions with relation to student scores, post-exam item analyses results and student perception. METHODS: This comparative cross-sectional study was conducted at the Dow University of Health Sciences, Karachi, from November 2010 to April 2011, and comprised medical students. Data was analysed using SPSS 18. RESULTS: Of the 207 participants, 16(7.7%) were boys and 191(92.3%) were girls. The mean score in Paper I was 18.62±4.7, while in Paper II it was 19.58±6.1. One-Best multiple-choice questions performed better than One-Correct. There was no statistically significant difference in the mean scores of the two papers or in the difficulty indices. Difficulty and discrimination indices correlated well in both papers. Cronbach's alpha of paper I was 0.584 and that of paper II was 0.696. Point-biserial values were better for paper II than for paper I. Most students expressed dissatisfaction with paper II. CONCLUSIONS: One-Best multiple-choice questions showed better scores, higher reliability, better item performance and correlation values.
OBJECTIVE: To find differences between One-Correct and One-Best multiple-choice questions with relation to student scores, post-exam item analyses results and student perception. METHODS: This comparative cross-sectional study was conducted at the Dow University of Health Sciences, Karachi, from November 2010 to April 2011, and comprised medical students. Data was analysed using SPSS 18. RESULTS: Of the 207 participants, 16(7.7%) were boys and 191(92.3%) were girls. The mean score in Paper I was 18.62±4.7, while in Paper II it was 19.58±6.1. One-Best multiple-choice questions performed better than One-Correct. There was no statistically significant difference in the mean scores of the two papers or in the difficulty indices. Difficulty and discrimination indices correlated well in both papers. Cronbach's alpha of paper I was 0.584 and that of paper II was 0.696. Point-biserial values were better for paper II than for paper I. Most students expressed dissatisfaction with paper II. CONCLUSIONS: One-Best multiple-choice questions showed better scores, higher reliability, better item performance and correlation values.
Entities:
Keywords:
One-Best MCQs, Type A MCQs, One-Correct MCQs, Reliability, Item analysis, Post-exam analysis, Student perception.